Great! Thanks again for taking the time. I'll give the patch a go. -Dominick
Matthew Flatt writes: > Thanks! I've pushed a repair. > > The problem was in the way that `raco distribute' updates an > executable's path to the Racket implementation. The path is in an > particular ELF section, and the section's size wasn't updated in the > header if the data was resized. > > I had trouble replicating the problem because, in my configuration, the > relevant path didn't get larger in the `raco distribute' step. > > At Wed, 24 Apr 2013 11:47:16 -0500, Dominick LoBraico wrote: >> Hi Matt, >> >> Thanks for your response! I'm sorry that I never got back to you, it looks >> like my gmail filters shuttled this away without my noticing. >> >> I'm attaching client.rkt for your reference. If you'd like the entire >> program, I could get that to you as well, just let me know. >> >> best, >> Dominick >> >> >> On Wed, Apr 10, 2013 at 8:34 AM, Matthew Flatt <[email protected]> wrote: >> >> > This problem looks familiar, but I think I wasn't able to replicate it >> > when I last tried, and I'm not able to replicate it now (even using >> > `funcitonal-command'). >> > >> > Could you send me "client.rkt", or some sharable variant that has the >> > same problem? >> > >> > Also, are you using a 32-bit or 64-bit build on Linux? >> > >> > Thanks! >> > >> > At Mon, 08 Apr 2013 17:00:29 -0500, "D. LoBraico" wrote: >> > > I have a command-line application that I have been compiling with Racket >> > > 5.3.3 as follows: >> > > >> > > raco exe client.rkt >> > > raco distribute client-dist client >> > > >> > > This creates a client-dist directory that I can then use to distribute >> > > my work to other machines running the same operating system. Up until >> > > this point I have been doing this on Mac OS X with no trouble. Now I am >> > > trying to replicate this process on a Linux machine. I run the same >> > > steps on a Linux box and have what appears to be the same client-dist >> > > directory. >> > > >> > > The difference, is that for some reason the version compiled on Linux >> > > does not work! That is to say, the binary in client-dist/bin does not >> > > function (even on the original machine that it was compiled on) in the >> > > same way as the binary that resulted from `raco exe >> > > client.rkt`. Specifically, it seems as if the executable thinks that it >> > > is getting more command-line arguments than it is. I modified my code to >> > > print all of the command-line arguments in hopes of getting some insight >> > > into this and get this: >> > > >> > > $ /tmp/client-dist/bin/client "test" >> > > client: expects no arguments on the command line, given 3 arguments: >> > "test" >> > > >> > > I suspect that whatever those 2 invisible arguments are is causing me >> > > trouble, but I can't for the life of me figure out where they are coming >> > > from. In case it is relevant, I am using the functional-command package >> > > from PLaneT for command-line parsing. >> > > >> > > Any ideas? >> > > >> > > best, >> > > Dominick >> > > ____________________ >> > > Racket Users list: >> > > http://lists.racket-lang.org/users >> > >> >> ------------------------------------------------------------------------------ >> [application/octet-stream "client.rkt"] [~/Desktop & open] [~/Temp & open] ____________________ Racket Users list: http://lists.racket-lang.org/users
